Cognition is thought to arise from the individual electrical interactions of large numbers of neurons that act in concert, yet much remains unknown about the dynamics of neuronal circuits in behavioral conditions. Recent technological advances permit a more thorough investigation of how 50-100 neurons collectively act in a microcircuit. In this collaborative proposal, a team of neurophysiologists, mathematicians and statisticians proposes to develop and apply state-of-the-art experimental and statistical tools towards the purpose of understanding the network structure of neuronal subpopulations in the prefrontal cortex of rodents. This work aims to shed light on connections between the functional dynamics of neuronal microcircuits and neuronal substrates of behavior, as well as provide fundamental tools for the study of microcircuits in systems neuroscience. The specific aims of the project are (1) the development of statistical models and accompanying analytical methods for inferring network properties of cortical microcircuitry from in vivo high-density multielectrod neuronal recordings in the presence of nonstationary background firing, (2) the use of simultaneous electrical recording and optogenetic stimulation techniques in order to validate and calibrate the statistical inferences, (3) the quantification of associations between network structure and neuronal representations of working memory, and (4) an investigation into the long-term plasticity of the network structure of prefrontal cortical subpopulations via precise optogenetic manipulation of their local dopaminergic inputs. Intellectual Merit Because of recent advances in brain recording technologies and the rise of optogenetic tools for direct manipulation of neuronal circuits, the field of systems neuroscience is poised for rapid advances in experimental findings and new understanding. This proposal sits at the forefront of this advance, combining state-of-the-art technology and statistical tools with novel experimental paradigms, brought together to elucidate the dynamics of neuronal networks in behaving animals. The proposed experiments have the potential to illuminate long-standing questions concerning neural substrates of learning and memory that have so far resisted direct experimental verification. The statistical tools under development have potential applications in diverse areas both within and outside of neuroscience. Broader Impacts: This proposal will fund the education and interdisciplinary training of graduate students and postdoctoral researchers, including the participation of underrepresented groups. It will fund the enhancement of research infrastructure via the collection and sharing of high-quality neurophysiological data, the development of publicly available statistical software, the advancement of experimental technologies and techniques, and the expansion and strengthening of collaborative interdisciplinary research partnerships. And it will likely spur the transfer of these neuroscientiic and statistical tools into other domains. The broad societal impact of neuroscience is of course extraordinary: extending influences to psychology, mental health, and medicine, as well as engineering and knowledge-based fields, even finally to philosophy and our most basic understanding of ourselves.